This section explores the relationship of the target country’s policy framework with the likelihood of cross-border deals, and with deal diversification. Sections 2 and 3 have highlighted the differences that exist between i) domestic and foreign M&A on the one hand, and ii) RECs and FITs on the other hand.
Econometric analysis will seek to relate the observed patterns of acquisitions – in particular, the relative likelihood of cross-border deals over domestic deals and the variation in diversification levels in different countries – to renewable energy policies adopted by these countries. Since the policy data are available only until 2011, the period is limited to 2005-2011. Furthermore, the analysis is restricted to the solar and wind sectors, because most of observations come from these two industries.16
Hypothesis 1: Since foreign acquirers face more information asymmetry and uncertainty, the response to policy characteristics (instrument choice, generosity) may differ between the two groups.
Hypothesis 2: Since FITs are set per technology and RECs for a basket of technologies, FITs are likely to be associated with a greater diversification of deals in terms of energy sources than RECs.
Both models are specified below, using letters F and G for the respective functions. Given the nature of dependent variables (a binary variable in the first equation, and a count variable in the second one), the first equation is estimated by Logistic regression and the second one by Poisson regression.17 Cross-country variation in FITs and RECs is explored to evaluate their relationship with the likelihood of cross-border acquisitions in the M&A sample and on the diversification of M&A deals. Deal diversification is defined as the number of new energy industries (out of 6) covered by the resulting entity.
Prob {Cross-border M&A =1} = F (FIT and REC generosity in the target country; Relative difference between the target and the acquirer in GDP, R&D intensity, investor protection; Deal characteristics;
sector and year fixed effects)
Deal diversification = G (FIT and REC generosity in the target country, sector and year fixed effects) The first regression includes additional control variables (Summary statistics can be found in the Appendix; Table 2 summarises the mean values of the main variables used in regression for domestic and cross-border deals). The most important deal characteristics are the technological motivation and the industry diversification. The technological motivation will be proxied by the dummy ‘Technological deal’, taking value 1 for deals involving a target and an acquirer, which both filed at least one application under the international Patent Cooperation Treaty in the last 10 years before the deal.18 The comparison of the mean value of this variable for domestic and crossborder deals suggests that foreign deals are more likely to be technologically motivated than domestic deals (see Table 2).
Furthermore, the regression includes a number of variables controlling for country pairwise differences. In particular, we control for differences in GDP, R&D intensity in each sector,19 and the level of investor protection. With respect to the first two variables, the literature shows that both relative market size and relative strength of a country’s science base drive FDI flows in research and development (Kuemmerle, 1999). Therefore, larger countries may be more attractive for cross-border entry into the renewable energy technology markets, and this is also what we see in our sample (Table 2). Furthermore, foreign countries with more research expenditures in a particular sector may have more innovative targets,
which may also attract foreign buyers. However, the opposite is also true, higher R&D intensity of the acquirer country is probably associated with more technologically advanced acquirers, who look for the possibility of locational diversification, which would lead to an opposite effect of pairwise differences in R&D intensity on the probability of a cross-border M&A. Therefore, the overall effect of this variable is generally ambiguous. For the cross-border deals that are present in our sample, target country has on average slightly lower R&D intensity than the acquirer country (Table 2).
Finally, the literature also suggests that the difference in corporate governance systems may be important for cross-border M&A (see Danbolt and Maciver, 2012 – on the analysis of particular aspects of governance systems). Therefore, we also control for differences in investors’ protection between countries, as acquirers’ probably prefer countries offering better protection to investors. To measure it, we use the
‘strength of investor protection’ index constructed by the World Bank (2013), which is the average of the extent of disclosure index, the extent of director liability index and the ease of shareholder suits index. The index ranges from 0 to 10, with higher values indicating more investor protection (Djankov, La Porta et al., 2008).20 Table 2 shows a greater value of investor protection in target country for cross-border deals.
For the sake of completeness, Table 2 includes also the policy variables that charactrise policy generosity in the target country. We observe a greater mean value of FIT policy and a lower mean value of REC policy for cross-border deals in comparison to domestic deals. Finally, we observe that crossborder deals are on average more diversified than domestic deals.
Table 2 Summary of the mean values of the variables for all, domestic, and cross-border deals
All deals Domestic
Difference in investor protection 0.192 0 0.490 449 289
Wind energy 0.478 0.468 0.492 449 311
FIT generosity in the target country 0.105 0.094 0.121 449 311
REC generosity in the target country 0.227 0.232 0.219 449 311
Diversification 1.771 1.653 1.942 449 311
Source: Own calculation using BNEF (2013), Renewable Energy Policy Database (OECD-EPAU 2013), IEA (2013a), World Bank (2013), and the OECD Statistics database.
The acquirer’s country policy may be seen as a control variable. However, the relationship to this control variable is ambiguous. On the one hand, a higher policy generosity level may weaken the incentive to enter another country market, because the acquirer already enjoys a generous policy support at home.
On the other hand, a possible mechanism is technology diffusion (e.g., Lovely and Popp, 2011;
Dechezleprêtre et al., 2012). According to this mechanism, acquirers with more advanced technologies may also be better capable of exploiting opportunities in less developed renewable markets, which offer other advantages (e.g., in terms of market size, less fierce competition, the presence of certain natural resources). Therefore the relationship is ambiguous. A robustness check has been conducted, showing that the inclusion of the acquirer’s policy generosity does not change the results.21
Finally, an important policy issue concerns a possible rent leakage to foreigners. Generous renewable energy policies generate rents which accrue to renewable energy producers. Therefore, governments might
undertake measures limiting foreign acquisitions of domestic targets to ensure that rents do not accrue to foreign firms. If so, only countries with strong protection against foreign M&As would have high FIT levels, which would cause endogeneity bias in our estimation. It is however unlikely that rent leakage was an important consideration in practice for two reasons. First, acquisition targets usually cover only a small market share (since targets are typically relatively small companies, and their number is small in comparison to the total number of companies benefiting from these policies). Second, the risk of rent leakage is mitigated by the fact that foreign acquirers need to compete with domestic acquirers that face lower transaction costs. Based on anecdotic evidence on the photovoltanic solar energy industry in Spain, foreigners did not seem to experience difficulties in this market during the high FIT period.22 Furthermore, countries in which stronger restrictions to incoming FDIs are observed (as measured by the OECD 2013 FDI Regulatory Restrictiveness Index) are generally also those with relatively less generous FITs.